Heat Mass Transfer (2017) 53:141–149

DOI 10.1007/s00231-016-1809-2

ORIGINAL

Upgrading of the thermal performance of two‑phase closed

thermosyphon (TPCT) using fusel oil
Adnan Sözen1 · Tayfun Menlik1 · Metin Gürü2 · Mustafa Aktaş1

Received: 17 March 2015 / Accepted: 29 March 2016 / Published online: 4 April 2016
© Springer-Verlag Berlin Heidelberg 2016

Abstract This study investigates how fusel oil affect the T Temperature (K)
thermal performance of a two-phase closed thermosyphon ΔT Temperature difference (K)
(TPCT) at various states of operation. The present study ρ Density (kg/m3)
experimentally demonstrated the effect of using fusel oil µ Viscosity (Pas)
comprised of various types of alcohols (1.1 % ethyl alco- η Thermal performance (−)
hol, 74.7 % amyl alcohol, 11.3 % isobutyl alcohol, 4.9 % S The function uncertainty
butyl alcohol and 3.8 % propyl alcohol and 4 % water) in WS The total uncertainty (%)
varying ratios on improving the performance of the TPCT. w1, w2, wn The uncertainties in the independent variables
Fusel oil has been obtained from fermentation plants as x Independent variables
a by product. A straight copper tube with an inner diam-
eter of 13 mm, outer diameter of 15 mm and length of 1 m Subscripts
was used as the TPCT. The fusel oil was filled up 33.3 % c Condenser
(44.2 ml) of the volume of the TPCT. Three heating power e Evaporator
levels (200, 300 and 400 W) were used in the experiments HP TPCT
with three different flow rates of cooling water (5, 7.5 and in Inner of the cooling water in the condenser
10 g/s) used in the condenser for cooling the system. An out Outer of the cooling water in the condenser
increase of 17.64 % was achieved in efficiency of TPCT w, water Deionized water
when fusel oil was used to replace deionized water at a heat
load of 200 W and with a cooling water flow rate of 10 g/s.
1 Introduction
Nomenclature
Cp Specific heat capacity (kJ/kg K) The most important aspects of heat transfer are energy-
k Thermal conductivity (W/m K) efficiency, speed and productivity. High thermal conduc-
ṁ Mass flow ratio (kg/s) tivity pipes fulfill these criteria, finding themselves in a
R Thermal resistance (K/W) wide range of applications. Their simple design, and other
Q̇ Heat transfer ratio (W) advantages including high rate of heat transfer, low cost,
low weight and low cost of maintenance have led to TPCTs
to becoming an attractive choice. A heat pipe consists of
* Adnan Sözen an insulated pipe and working fluid and additionally a wick
[email protected] in cases where gravity is not utilized in fluid transfer. The
http://websitem.gazi.edu.tr/asozen
pipe is comprised of three major parts: lower evapora-
1
Energy Systems Engineering, Technology Faculty, Gazi tor part, the adiabatic region in the central range and the
University, 06500 Teknikokullar, Ankara, Turkey condenser at the top. The choice of working fluid is critical
2
Chemical Engineering, Engineering Faculty, Gazi University, for the optimization of the thermal performance due to the
06570 Maltepe, Ankara, Turkey fact that it utilizes the phase change of the working fluid

13
142 Heat Mass Transfer (2017) 53:141–149

for heat transport. Recent studies have focused on the use in fusel oil constitute the ideal mixture with which to
of metal oxide containing nano-fluids as the working fluid obtain high performance in TPCTs. The subject study has
for improved TPCT performance. Nanofluids have been acknowledged this fact with the design of an experimental
successfully applied as the working fluid in TPCTs for the setup to investigate the performance of TPCTs when fusel
improvement of the heat transfer performance owing to the oil is used.
higher thermal conductivity of solids than that of liquids
[1–10]. Additionally, alcohol-based solutions have been
used to improve the performance of heat pipes [11–13]. 2 Experiments
Fusel oil, a byproduct of the distillation step of the fer-
mentative ethyl alcohol production industry, is a source 2.1 Experimental setup
of amyl alcohols. The composition and concentration of
fusel oil varies, depending on the source of carbon used in The experimental setup is schematically represented in
the process of fermentative alcohol production and on the Fig. 1 and also photograph of the experimental set up is
method employed for the separation of the fusel oil from given Fig. 2. A straight copper tube with an inner diam-
the fermentation mixture. Fusel oil is comprised mainly eter of 13 mm, an outer diameter of 15 mm and a length
of low-molecular-weight alcohols (mainly i-amyl alco- of 1 m, of which 400 mm was the evaporator, 200 mm
hol, i-butyl alcohol, n-propyl alcohol, n-butyl alcohol, was the adiabatic section and the remaining 400 mm was
ethyl alcohol and n-amyl alcohol), some water, and trace the condenser, was used as the TPCT in the present study.
amounts of aldehydes, free acids and their esters, higher The evaporator region was heated using an electrical heater
alcohols and terpenes. As is shown in Table 1, ethyl alco- with a nominal power of 1500 W, which was comprised
hol, butyl alcohol and propanol (propyl alcohol) are the of a Ni–Cr wire. The heater was fully insulated. The input
chief constituents of fusel oil, but it also contains a small power was measured and monitored through the use of a
amount of water. Therefore, it is referred to as a waste alco- wattmeter (Fluke-43b analyzer). Power was experimen-
hol mixture due to its impurity. Additionally, alcohols have tally measured in terms of Watt. The heater (the evapora-
more polarity than water. As a result of increased polarity- tor section) was completely insulated and therefore the
due mainly to the presence of water–fusel oil is highly heater power measured in the wattmeter is assumed to be
effective in solving inorganic oxide compounds. There is completely transferred to the study fluid. Based on the fact
no economic value of fusel oil other a handful of common- that 100 % isolation is impossible, the heat pipe was iso-
place applications. However, we thought it could be used as lated using 8 cm fiberglass, the heat loss was calculated for
a fluid in a cooling system due to its multilevel volatility. each evaporator heat input (200, 300 and 400 W), and 10 %
Fusel oil containing ethyl (1.1 % wt/wt), amyl (74.7 %), of heat loss to the surroundings was determined based on
isobutyl (11.3 %), butyl (4.9 %), propyl alcohol (3.8 %) and available literature data. This amount of heat loss was con-
water (4.1 %) was used as the working fluid in the TPCT of sidered during the calculation of thermal resistances and
the present study. The ethyl and propyl alcohol in fusel oil efficiencies of TPCT.
lower the evaporation point due to their low carbon content. The condenser section of the TPCT was furbished with
Meanwhile, the increase in the relative composition of the a cooling chamber. The heat was removed from the con-
isomers among the alcohols with the same number of car- denser via the circulation of the coolant in the chamber. A
bon atoms (isobutyl alcohol) also lowers the boiling point mass flow meter was used to control and set the flow rate
from the normal temperature. Furthermore, amyl alcohol of the coolant as required. Glass wool was used to wrap all
facilitates condensation. The working fluid in a TPCT is sections of the TPCT in order to allow minimal loss of heat
expected to facilitate evaporation at a low temperature and from the system.
condensation at a high temperature. The alcohols present Ten thermocouples (k-type) were used to monitor the
distribution of the temperature along the wall of the TPCT.
Four of these were mounted on the evaporator and four
Table 1  Chemical composition of fusel oil
others were placed on the condenser section whereas the
Alcohol % (wt/wt) Chemical composition remaining two were placed on the inlet and the outlet of the
Ethyl alcohol 1.1 C2H5OH
coolant to monitor the temperature change of the coolant.
Figure 3 indicates the locations of the thermocouples along
Amyl alcohol 74.7 C5H11OH
the TPCT. The details of the placement of thermocouples
Isobutyl alcohol 11.3 C4H9OH
for measurement of the inside surface of the TPCT were
Butyl alcohol 4.9 C4H9OH
shown in Fig. 3 for evaporator and condenser separately.
Propyl alcohol 3.8 C3H7OH
The recordings of the thermocouples were monitored with
Water 4.1 H2O
a data-logger (E-680, Elimko, Turkey).

13
Heat Mass Transfer (2017) 53:141–149 143

Fig. 1  Schematic view of the

experimental set-up

The working volume of the fluid was kept the same

(44.2 ml) when either deionized water or the fusel oil was
used in the system. This value corresponds approximately
to 1/3 of the total volume of the TPCT. The TPCT was
positioned a 90° angle to the horizontal. Input power values
of 200–400 W with increments of 100 W were utilized. The
flow rate of the cooling water, which was used to condense
the working fluid in the condenser, was set to three differ-
ent values; 5, 7.5 and 10 g/s.
It can be seen in the experimental results (Sect. 3) that
saturation temperature (65 °C) was reached at vacuum
pressure (25 kPa) when deionized water was used as the
working fluid. This result indicates that saturation condi-
tions were reached inside the TPCT. At the same pressure
of constant composition mixture, the pressure drop was
negligible.

2.2 Uncertainty analysis

In the experiments, the data are obtained with appropri-

Fig. 2  Photograph of the experimental set up
ate measurement instruments. However, uncertainties may

13
144 Heat Mass Transfer (2017) 53:141–149

Fig. 3  Configuration and loca- Tout

tion of all thermocouples along

12 cm 12 cm 12 cm 4 cm
the TPCT
Tc4

Tc3
Condenser
Thermocouple
Tc2

Tc1
Detail of thermocouple connection

20 cm
Tin
Adiabatic

Te4

12 cm
Te3

12 cm
Thermocouple
Evaporator
Te2 12 cm

Te1 Detail of thermocouple connection

4 cm

arise. Errors and uncertainties can arise from instrument

selection, test conditions, calibration, environment, obser- 2 1/2
 2  2 
WT = wthermocouple + wconnection points + wreading
vation, reading, test planning and connection type-points  1/2
of measurement probes. Therefore, uncertainty analysis is = (0.5)2 + (1)2 + (0.5)2 = 1.224
important to obtain the desired experimental standards. The
result S is a given function in terms of the independent var- The total uncertainty in the measurement of the flow (ṁ)
iables x1, x2, …, xn. Let Ws be the uncertainty in the result can result from flowmeter and reading. This uncertainty
and w1, w2, …, wn be the uncertainties in the independent may be calculated as follows:
variables. If the uncertainties in the independent variables  2  2 1/2
are all given with the same odds, then the uncertainty in the Wm = wflowmeter + wreading
result having these odds is given as [14].  1/2
= (0.01)2 + (0.01)2 = 0.0141
 2  2  2 1/2
∂S ∂S ∂S
WS = w1 + w2 + ··· + wn The total uncertainty in the measurement of the power of
∂x1 ∂x2 ∂xn
(1) heater (Ph) can result from Wattmeter analyzer and reading.
This uncertainty may be calculated as follows:
Uncertainty analysis is a powerful method, especially  2  2 1/2
when it is used in planning and design of experimental WPh = wWattmeteranalyser + wreading
studies.  1/2
The temperature, flow and heater power variables were = (1)2 + (1)2 = 1.414
measured during the experiments. The total uncertainty in
the measurement of the temperature (T) can result from When the uncertainties of measured variables (tempera-
thermocouple (wthermocouple: accuracy error), connection ture, flow, and heater power) were analyzed, it was deter-
points (wconnection points: error caused by fittings and connec- mined that these values are in an acceptable range. Tech-
tion points of probes) and reading (wreading: error caused by nical properties of measurement equipment, accuracy, and
percentage parts of measured data for digital instrument). total uncertainty are shown in Table 2.
This uncertainty may be calculated as follows:

13
Heat Mass Transfer (2017) 53:141–149 145

Table 2  Technical properties of measurement equipments, accuracy, total uncertainty

Devices Technical properties Accuracy Total uncertainty

Thermocouples (data logger, Elimko E-680 model) K type; measurement range −200 to +1200 °C ± % 0.5 °C ± % 1.224 °C
Flowmeter Measurement range 0–12 g/s ± 0.01 g/s ± 0.0141 g/s
Wattmeter Fluke-43b analyzer ±1W ± 1.414 W

Fig. 4  The viscosity of fusel oil in comparison with deionized water Fig. 5  Thermal conductivity of fusel oil in comparison with deion-
ized water

2.3 Thermophysical properties of fusel oil water prior to conducting the measurements for fusel oil.
The results obtained in the analysis are shown in Fig. 5.
The relevant thermophysical properties of fusel oil were The thermal conductivity of fusel oil is, on average, 13 %
determined before the experiments were conducted. The higher than that of water.
density was calculated by weighing a known volume of
fusel oil. The specific heat capacity was measured using
differential scanning calorimetry. The temperature of the 3 Results and discussions
water and fusel oil sample increased at the same rate as a
reference substance, and the specific heat of the water and The inlet and the outlet cooling water temperature differ-
fusel oil sample was calculated by measuring the differ- ence in condenser, the water mass flow rate and the specific
ence in heat required to raise the temperature by a known heat were used to calculate the quantity of heat transferred
amount. The specific heat of fusel oil is 4.36 kJ/kg K. to the coolant water in the following way:
A rheometer (model DV-III Ultra) and a water bath
(model TC-502) from Brookfield were used in conducting
Q̇c = ṁc · cp · (Tout − Tin ) (2)
the viscosity measurements. A spindle (No. SC4-21) was The ratio of the output heat by condensation to the inlet
used in the measurements. The device was calibrated using heat by evaporation is described as the efficiency of the
pure water at ambient temperature, and the viscosities of TPCT and is given as:
the 50 ml fluid samples were measured at 20°, 40°, 60°
and 80 °C of temperature. The samples were warmed to Q̇c Q̇out
the desired temperature in an oil bath prior to analysis. The
η= = (3)
Q̇e Q̇in
analysis of the 50 ml sample with undetermined viscosity
was conducted following this warming-up period through where Q̇e is the heating power (200, 300 and 400 W).
the use of a spindle rotating at a constant speed of 50 rpm The thermal resistance between the evaporator and the
at constant temperature. The data collected from the analy- condenser section is defined as:
sis is shown in Fig. 4 as a plot of the variation in viscosity.
T
The coefficient of thermal conductivity of fusel oil R= (4)
Q̇in
was determined in an experimental setup in compliance
with ASTM Standard D2717-95, “Standard Test Method where Q̇in = Q̇e is the heating power (200, 300 and 400 W)
for Thermal Conductivity of Liquids.” The measure- ΔT is the temperature difference between mean tempera-
ments made by the experimental setup were verified for ture of the evaporation (the average of the temperatures

13
146 Heat Mass Transfer (2017) 53:141–149

Fig. 6  Distribution of wall temperatures versus different mass flow

rates of coolant. a 5 g/s, b 7.5 g/s, c 10 g/s in condenser of TPCT for
Fig. 7  Distribution of wall temperatures versus different mass flow
200 W input power
rates of coolant. a 5 g/s, b 7.5 g/s, c 10 g/s in condenser of TPCT for
300 W input power

measured at four locations in the evaporator) and mean into the system, respectively. Captions a, b and c in the
temperature of the condensation section. ΔT is given by: figures indicate the outer wall temperature with varying
    flow rates of cooling water (5, 7.5 and 10 g/s) allowing
Te1 + Te2 + Te3 + Te4 Tc1 + Tc2 + Tc3 + Tc4
T = − the heat transfer in the condenser and condensation of the
4 4
working fluid. Generally speaking, a gradual decrease was
(5) observed in the temperature of the pipe wall in the direc-
The heat transfer of a TPCT uses a working fluid that tion from the evaporator to the condenser. An example
changes phases in a continuous endothermic and exother- of this observation would be the case with a power input
mic cycle, giving the TPCT excellent heat transfer perfor- of 200 W and condenser cooling water flow rate of 5 g/s
mance [15]. (Fig. 6a), in which theaverage wall temperature of evapo-
Improvement of the efficiency of the TPCT was deter- rator Te1 +Te2 +T e3 +Te4
was shown to decrease from 68.9
4
mined using deionized water and fusel oil. The tests were to 56.9 °C upon the replacement of deionized water with
conducted initially by charging the TPCT with deionized the fusel oil. Thisindicated an average decrease  of 12 °C
water and then with the fusel oil.

Te1 +Te2 +Te3 +Te4 Te1 +Te2 +Te3 +Te4
4 − 4
Figures 6a–c, 7a–c and 8a–c display the distribution fusel-oli dionizedwater
of the wall temperature along the TPCT, which was posi- on the average wall temperature of the evaporator (Te).
tioned vertically, with deionized water or the fusel oil as the The highest difference in the temperature of the wall
working fluid using 200, 300 or 400 W of power supplied was recorded as 19.4 °C under similar conditions of input

13
Heat Mass Transfer (2017) 53:141–149 147

Fig. 9  Total thermal resistance of TPCT for varying flow rates of

coolant in the condenser having heat loads of a 200 W, b 300 W and
c 400 W
Fig. 8  Distribution of wall temperatures versus different mass flow
rates of coolant. a 5 g/s, b 7.5 g/s, c 10 g/s in condenser of TPCT for absorbs more heat in order to transform into vapor phase
400 W input power
at working pressure. Therefore, it is able transform into
vapor phase at temperatures higher than water tempera-
power of 200 W and cooling water flow rate of 10 g/s with ture at higher heat levels. Condensation rate is higher at
the fusel oil (Fig. 6c). condenser zone due to amount and physical properties of
Since decreasing the temperature difference ΔT using amyl alcohol. Thus, ΔT shrinks and decreases thermal
fusel oil causes a decrease in thermal resistance, the evapo- resistance due to the property of vaporization at low tem-
ration of the fluid in the evaporator section at lower tem- peratures and condensation of high temperatures that is
peratures and the rapid transfer of heat to the cooling fluid expected from heat pipes.
in the condenser section are caused by its high thermal In experimental study, vaporization has been occured by
conductivity. Therefore, the improvement of both endother- means of heater. As a result of boiling of fusel oil; ethanol,
mic and exothermic heat transfer would result in the use of propanol, water, isobutanol, butanol and amyl alcohol are
fusel oil as more advantageous fluids over water. vaporized fractionally from the fluid, related to the vola-
Fusel oil is a waste product that is composed of differ- tility. Vapor has been recondensed absolutely by means of
ent alcohols. Ethyl and propyl alcohols within fusel oil cooling system in condenser. There is no vapor at the top of
increase the evaporation rate with increasing heat input experimental set-up. Therefore, it is not necessary devola-
through the evaporator zone; however, due to the fact that tilization process in our set-up. Since working fluid (fusel
boiling point of 74 % amyl alcohol in its composition is oil) does not content non-condensable gases, there are no
higher than water (BP: 138 °C at atmospheric pressure), it gasification and degassing processes.

13
148 Heat Mass Transfer (2017) 53:141–149

rapidly to a minimum with increasing heat loads as dis-

played in Fig. 9a–c. This finding is also in accordance with
the other results, indicating that the thermal resistance
(R) of the TPCT improved when the fusel oil was used in
place of water as the working fluid (Fig. 9). A decrease of
35.96 % was achieved in thermal resistance (R) when fusel
oil was used to replace deionized water at a heat load of
200 W and with a cooling water flow rate of 10 g/s. The
average improvement is approximately 22 % for other heat
loads and cooling flow rates.
Figure 10 represent the efficiency and the improve-
ment in efficiency with respect to different mass flow rates
of coolant in the condenser provided with varying power
inputs of 200, 300 and 400 W. A increase of 17.64 % was
achieved in efficiency of TPCT when fusel oil was used to
replace deionized water at a heat load of 200 W and with a
cooling water flow rate of 10 g/s. The increase in the effi-
ciency of the TPCT for other heat loads and cooling flow
rates are presented in Fig. 10. The comparison of the aver-
age of 7.2 % improvement in the efficiency with the 22 %
decrease in thermal resistance in Fig. 10 indicates that the
fusel oil cannot exhibit the same capability it has in heat
absorption in the endothermic phase change for the transfer
of the heat it absorbed to the cooling water in the exother-
mic phase change.

4 Conclusion

The experimental system in this study comprised of a

TPCT charged with fusel oil suspended in deionized
water and the experimental results were compared with
those of a TPCT that was charged with deionized water as
the working fluid. The results of this study on the evalu-
ation fo the performance of a TPCT led to the following
Fig. 10  Efficiency of the TPCT for varying flow rates of coolant in
the condenser having varying heat loads (200 W, 300 W and 400 W) conclusions:

1. The distribution of the wall temperature was lim-

In this study, three different evaporation heats were ited to a narrower range when the fusel oil was used
used in order to optimize the TCPT working tempera- rather than deionized water. Similarly, the difference
ture interval. Three different cooling water flow rates between the wall temperature of the evaporator and
were used in order to optimize the collected heat at the condenser was smaller using the fusel oil as the
the condensation region. Therefore, analysis results working fluid than when deionized water were utilized
have been obtained for nine different cases (Figs. 6, 7, as the working fluid.
8) and the changes in thermal resistance and heat pipe 2. The evaporation temperature of the fusel oil was deter-
efficiency are given at three different points for both mined to be 12 °C lower in the evaporator than that
heat and flow rate in accordance with these parameters of deionized water. This result shows that the transport
(Figs. 9, 10). of heat from the low temperature hot sink would be
Figure 9a–c represent the thermal response implicated allowed along the TPCT.
by the R versus Q plots of water and the fusel oil at varying 3. The thermal resistance was lowered by 35.96 %
coolant flow rates in the condenser of the TPCT. The ther- through the use of fusel oil, respectively.
mal resistance was determined to be high for both water 4. The thermal performance of the fusel oil was better
and the fusel oil at low heat loads although it decreases (average 17.64 %) than that of deionized water.

13
Heat Mass Transfer (2017) 53:141–149 149

References 8. Naphon P, Assadamongkol P, Borirak T (2008) Experimental

investigation of titanium nanofluids on the TPCT thermal effi-
ciency. Int Commun Heat Mass Transf 35:1316–1319
1. Saleh R, Putra N, Prakoso SP, Septiadi WN (2013) Experimental
9. Sözen A, Menlik T, Gürü M, Öztaş S (2015) Heat transfer
investigation of thermal conductivity and TPCT thermal perfor-
enhancement using MgO/water nanofluid in TPCT. J Energy Inst
mance of ZnO nanofluids. Int J Therm Sci 63:125–132
88:247–257
2. Sözen A, Menlik T, Gürü M, Irmak AF, Kılıç F, Aktaş M (2016)
10. Noie SH, Heris SZ, Kahani M, Nowee SM (2009) Heat transfer
Utilization of fly-ash nanofluids in two-phase closed thermosy-
enhancement using Al2O3/water nanofluid in a two-phase closed
phon (TPCT) for enhancing heat transfer. Exp Heat Transf
thermosyphon. Int J Heat Fluid Flow 30:700–705
96:330–337
11. Zhang XM, Xu JL, Zhou ZQ (2004) Experimental study of a
3. Shanbedi M, Heris SZ, Baniadam M, Amiri A (2013) The effect
pulsating TPCT using FC-72, ethanol, and water as working flu-
of multi-walled carbon nanotube/water nanofluid on thermal per-
ids. Exp Heat Transf 17(1):47–67
formance of a two-phase closed thermosyphon. Exp Heat Transf
12. Francescantonio N, Savino R, Abe Y (2008) New alcohol solu-
26(1):26–40
tions for TPCTs: Marangoni effect and heat transfer enhance-
4. Wei WC, Tsai SH, Yang SY, Kang SW (2005) Effect of nano-
ment. J Heat Mass Transf 51:6199–6207
fluid concentration on TPCT thermal performance. IASME
13. Savino R, Abe Y, Fortezza R (2008) Comparative study of
Trans 2:1432–1439
TPCTs with different working fluids under normal gravity and
5. Naphon P, Thongkum D, Assadamongkol P (2009) TPCT effi-
microgravity conditions. Acta Astronaut 63:24–34
ciency enhancement with refrigerant-nanoparticles mixtures.
14. Akpinar EK (2010) Drying of mint leaves in a solar dryer and
Energy Convers Manag 50:772–776
under open sun: modelling, performance analyses. Energy Con-
6. Hajian R, Layeghi M, Sani KA (2012) Experimental study of
vers Manag 51:2407–2418
nanofluid effects on the thermal performance with response time
15. Kang S-W, Wei WC, Tsai SH, Huang CC (2009) Experimental
of TPCT. Energy Convers Manag 56:63–68
investigation of nanofluids on sintered heat pipe thermal perfor-
7. Khandekar S, Joshi YM, Mehta B (2008) Thermal performance
mance. Appl Therm Eng 29(5–6):973–979
of closed two-phase thermosyphon using nanofluids. Int J Therm
Sci 47:659–667

13